Dr. Ciprian Crainiceanu obtained his PhD in statistics from Cornell University in 2003 and is currenly a Professor of Biostatistics at Johns Hopkins University. He is a generalist who specializes in wearable and implantable technology (WIT) in health studies and Neuroimaging in clinical studies. He is the co-founder of the Statistical Methods and Applications for Research in Technology (SMART, www-smart-stats.org) research group and the co-founder of Neuroconductor (www.neuroconductor.org).
Talk: Functional Data Analysis with R
Abstract: Functional Data Analysis (FDA) provides a conceptual framework for analyzing functions instead of or in addition to scalar measurements. For example, physical activity is a continuous process over the course of the day and can be observed for each individual; FDA considers the complete physical activity trajectory in the analysis instead of reducing it to a single scalar summary, such as the total daily activity. In this talk I will show how nonparametric smoothing methods can be extended to functional regression using the mixed effects model framework. We will illustrate the impact of these ideas on the practical implementation of these methods in the R package refund. Methods are motivated by and applied to the association between objectively measured physical activity using body-worn accelerometers and health outcomes in the National Health and Nutrition Examination Survey (NHANES) and the UK Biobank studies.
Bio